Modifiable and non-modifiable factors associated with functional impairment during the inter-episodic periods of bipolar disorder
BACKGROUND Growing interest has been given to the construct of Duration of untreated illness (DUI) on the outcome of bipolar disorder (BD), due to its potentially modifiable nature. The aim of this study was to identify possible clinical correlates of DUI in a sample of BD patients. METHOD 119 BD spectrum patients included. DUI rate was calculated and dichotomized into short DUI and long DUI subgroups, cut-off 24 months. These subgroups were compared for socio-demographic and clinical variables. Significant results were included into direct logistic regressions to assess their impact on the likelihood of presenting with long DUI. RESULTS Mean DUI±SD was 75.6±98.3 months. Short DUI subgroup comprised 56 (47.1%), long DUI 60 (52.9%) patients. Age at onset of BD was lower in the long DUI subgroup (p=0.021), illness duration longer (p=0.011). Long DUI subgroup showed significantly more comorbidity with Axis I (p=0.002) and personality disorders (p=0.017), less interepisodic recovery (p<0.001) and less Manic Predominant Polarity (p=0.009). Direct logistic regression as a full model was significant, correctly classifying 76.7% of cases. A unique statistically significant contribution was made by: Manic Predominant Polarity, Personality Disorder Comorbidity, and Total Changes in Medications. LIMITATIONS Partial retrospective data, cross sectional study. CONCLUSIONS DUI was longer than 24 months in half of the sample. Psychotic /Manic onset contributed to a quick diagnostic classification. Personality disorders in depressed patients could delay a correct diagnosis of BD, factors associated with an increased likelihood of BD must be considered. More research on personality disorder comorbidities is needed.